AMBER Archive (2005)

Subject: Re: AMBER: problem with input files when using ff02EP + pol3

From: David A. Case (case_at_scripps.edu)
Date: Mon Apr 04 2005 - 00:18:29 CDT


On Sun, Apr 03, 2005, Phineus Markwick wrote:
>
> On loading the pdb file I am told that numerous 'atoms' are missing...
> these turn out to be the 1EP+ and 1EP- point charges that mimic the presence
> of the electron lone pairs. On the check command I get a large number of
> warnings of close contacts between the point charges.

This is all OK: the "extra point atoms" are indeed missing from your input pdb
file, but LEaP will figure out where to put them. The close contact warnings
were never updated to take into account extra points -- you can ignore these.

>
> getting new box info from bottom of inpcrd
> | peek_ewald_inpcrd: Box info not found in inpcrd

The default in Amber 7 and later is ntb=1, which implies a periodic
simulation. But it looks like you have a non-periodic simulation, and hence
no "box info", as implied by the above error message.

Note that your "second tleap script" did not have a solvatebox command; if
this file has all of the waters, you still would need to issue a "setBox"
command to get the system to be a periodic one. By far the easiest thing to
do is just to use a single tleap script, that loads the protein, does
solvateBox and addIons, and saves the resulting (periodic) prmtop file.

(Basically, the first script you posted created a periodic system, but when
that was saved and re-loaded as a pdb file, the periodic information was lost,
since PDB files have no place to store this information. So the second time
around, LEaP created a non-periodic system.)

...good luck...dac

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu